[John3]

Our Swiss slant with Fanuc 0T-B is driving us crazy.

When we turn the machine on to setup for and run a new program it acts very inconsistently in the X-axis.

Sometime the X-axis gets lost (by 7+") when doing MDI G68/G69 mirroring for the front and rear turrets.

Yesterday we touched off all the tools, entered the offsets and ran 250 pieces without a single problem.

This morning we turned the machine on to start running more of the same part. The first part was whacked out big time. When we made test cuts we found that ALL X-axis offests were off somewhere between .100 and .160. Fortunately we didn't crash anything or break any tools.

After resetting the offsets (to different numbers from yesterday) the machine is again making great parts, not a single hick-up. So far its at part 189 for the day.

What could be causing this? It doesn't seem to be servo or mechanical problems or it would show up in the parts wouldn't it?

John

[bob1371]

If you have limit switches for the zero return check to ensure they are not loose, that the rollers don't have flat spots and the dogs are ok. It could be homing to a different position during th initial zero return.

Bob

[cncdiag]

Sounds like your zero return position is changing.
Does it change by the pitch of the ballscrew?
Make sure the decel switch for zero return is not sticking.
If the switch is fine, try moving the dog for the decel switch a little.

[John3]

Sounds like your zero return position is changing.
Does it change by the pitch of the ballscrew?
Make sure the decel switch for zero return is not sticking.
If the switch is fine, try moving the dog for the decel switch a little.

Thanks for the ideas guys.....

It DOES NOT seem to be off by the pitch of the ball screw!!

It is off by something a little less than that.

One rev of the ball screw is 10.000 machine counts, which because its X (diameter) means the ball screw pitch is 5mm or 0.1969". The biggest adjustment that we've made to tools to get part sizes correct is .162".

The ball screw is a direct drive off of the a red-cap motor. Everything seems tight.

I will reposition the decel switch a bit to see if it changes things.

Could there be some sort of bug / or race condition in the routine that does the G68 mirror function? I wounder if that is doing some sort of grid shift or G50 Reference Register behind the scenes? Is there any way I could look into these theories if they have any merit?

John